The 1996 pertussis epidemic in New Zealand: vaccine effectiveness

Aim. To assess if reduced vaccine effectiveness-may have accounted for incr eased hospitalisations in the 1996 pertussis epidemic. Methods. Vaccine effectiveness was estimated by comparing vaccine coverage of the population (derived from a literature review) with that of cases (fr om notification data available from 1 June 1996) - the screening method. On ly three doses of pertussis vaccine were in the immunisation schedule until 1996, so vaccine effectiveness was calculated for three or more doses. Results. Most likely estimates of vaccine effectiveness for Europeans were 88% (95% confidence interval 71 to 95%) for 5- to 14-month-olds, 80% for 15 -month to 4-year-olds (66 to 88%) but lower for children aged 5 years and o lder with confidence limits including zero. Vaccine effectiveness estimates for Maori were less for each age group but based on few observations. Conclusions. The increase in hospitalisations for young children in the 199 6 epidemic cannot be directly attributed to a reduced vaccine effectiveness , as vaccine effectiveness estimates for preschool Europeans are in line wi th international evidence. Additionally, the vaccine effectiveness estimate s in this study are likely to be underestimates due to bias. The lower esti mates for vaccine effectiveness among Maori are likely to reflect increased pressure of these biases, although a biological basis for the difference o r clustering of factors that cause failure are also possible. The vaccine e ffectiveness estimates decrease with age, a likely combination of waning va ccine immunity and the cross-sectional nature of the screening method itsel f for determining vaccine effectiveness.